It’s Friday, the end of the work week, and all everyone wants to do is get home as quickly as possible. For the transit rider, it is time to enter the arena of unknown bus reliability. Will my bus come? Will it be on time? How bad will traffic be? We have all mentally asked these questions, but some have to ask them more than others.

Today we will look at the 5 buses that have the worst afternoon reliability in the Metro system and consider what can be done to improve them. These routes are generally low ridership and wouldn’t merit much capital investment, so we’ll focus on quick fixes where appropriate. Conveniently, each one of our tardy routes is from a different portion of the county. (On time data is from the King County Metro 2018 System Evaluation. )

Route 63

The North Seattle bus route is late 42% of the time during the afternoon commute. Route 63 is a 9 mile peak express that only runs on weekdays and carries 700 passengers per day.

As you can see from the map above, the route is complex with different alignments in each direction. The most notable difference is the that the bus uses the express lanes in the morning to access Mercer Street, but uses the mainline to go north in the afternoon while providing freeway stop service to the U-District. It is a virtual certainty that the route 63 will be revised whenever Metro does the restructure that will be related to the opening of Northgate Link. Investments like signal timing and bus lanes would probably not become active in time to help.

So here are two changes that could occur that might help the riders of the 63 between now and 2021. First, eliminate the diversion over to 8th Avenue in the afternoon and keep the bus on Boren. Riders in the afternoon on 8th should be able to find other northbound options give how close they are to Westlake Center. Second, eliminate the afternoon freeway stop service to the U-District. Google Maps provides multiple alternative routings to the U-District all of which offer more convenience and freedom of choice than being dropped off on 45th street next to I-5. These two actions should help improve reliability until the expected northeast Seattle restructure occurring sometime in 2021.   

Route 179

Next on the docket of unreliable bus lines is the 179 with 47% of the afternoon buses run late. The bus route is 28 miles long and carries 800 passengers per day into Seattle.

The 179 is the epitome of a peak express. It wanders through neighborhoods, picking up people and hitting park and rides, then jumps on the freeway to Seattle. In the short run, adding bus signal priority in the three locations marked on the map would help improve reliability to a modest degree. In the future, adjustments will almost certainly happen to the 179 when Federal Way Link opens in 2024. The most likely outcome is that the 179 is truncated at Federal Way Transit Center in exchange for all day service. 

Route 37

Our Seattle based example of poor reliability is the 37 which runs late 49% of the time in the afternoon. The 37 is a milk run that starts at the Alaska Junction and loops around the Alki Peninsula before working its way to the Seattle core. It carries about 200 passengers per day and is 14 miles long.

In the short run, signal priority and bus lanescould be used to help improve this run. But given the coverage purpose of the 37, I am not sure how much they will help. In the very long run, we can expect these buses to be truncated at the West Seattle light rail. However, with Link service not starting until 2030 this route, which provides coverage to the Alki neighborhood, needs to improve its punctuality and passenger count.

Therefore, I would propose to split the route into 37N(orth) and 37S(outh). The schism would occur at 63rd & Admiral Way with 37S covering the route to a transfer to the C-Line at the Alaska Junction and on to 35th Avenue. Stops should be added at Totem Pole Park to allow a transfer to the 21. This would prepare the 37S for a future transfer to Link at Alaska Junction.

Service north of 63rd & Admiral Way would be provided by the 37N. Transfers could happen on Spokane Street to the 21 and the route would be extended slightly to end at SW Andover Street and Delridge Way to connect with the 120 and the future Delridge Link Station. If that is too complex, the bus should end at Sodo station to avoid downtown traffic.

Routes 216/219

The 216 and 219 are the last routes to review and they represent east King County. The 216 starts at the Bear Creek Park and Ride and proceeds with stops all the way to the Issaquah Highlands Park and ride before converting to an express to downtown Seattle. The 219 has a similar rooting but starts at WA SR 202. Combined, the 216 and 219 have 30 mile long routes, carry 1700 passengers per day , and are late 55% and 54% of the time, respectively. Those reliability figures are the worst of all of Metro’s routes.

The biggest challenge with fixing these two routes is that they are long and have congestion throughout. The most aggressive solution would be to cancel both. All but one of their stops are currently served by the 269 that runs between the Issaquah Transit Center and Overlake. The service hours saved could be used to increase the 269 to at least 15 minute service. It would also allow for better connections to more destinations and make service for Sammamish less Seattle-centric. However, such an action would be very aggressive with East Link opening in only a few years. Perhaps it’s best just to wait for the East Link restructure.

In the Seattle area, the afternoon peak tends to be more congested than the morning peak. The buses listed above have a few things in common: they are peak expresses, run only on weekdays, and the majority of them are long. All of these factors lead to a lack of reliability.  Fortunately, grade separated light rail will to allow the truncation and restructuring that should allow more reliable PM commutes.

108 Replies to “Metro’s least reliable routes”

  1. Interesting. All of these are peak expresses, and 4/5 of them are peak from downtown Seattle. When I went to Seattle a lot, the 2nd Ave southbound buses in the evening all had big reliability problems. I took mainly the 158/159 on Friday afternoons, and it was maddeningly unreliable. Whatever space Metro has downtown to keep these afternoon trippers should be expanded, and most of these would probably get a lot better.

    I used to take the 179 a lot when I went to Seattle from Federal Way. I would usually walk farther to the 179 stop (rather than take a bus to the TC and transfer to the 577) because it’s under-utilized before the transit center, so I’d usually score a seat on the bus and be able to take a nap all the way to Seattle.

    Once Link opens, there’s no point in keeping the standalone 179 since the 181 west portion has a nearly identical route. It works nicely right now since they want both more peak frequency on 320th and extra peak capacity to Seattle, so it makes some sense to combine them into a new peak express route.

    1. I disagree about routes 158/159. They are taking up space for other routes that don’t have Sounder as a (usually) reliable alternative. Sounder is nowhere near crushloaded, and people who need a seat are able to get one if they ask. Same goes for 157, except coming out of Tukwila Sounder Station. Then, invest the saved service hours into more and more frequent, and two-way, local connections to the stations, better-timed to the train schedules.

      Downtown removals also ought to include 101/102, coming out of Rainier Beach Station, creating more connectivity. Ideally, the I-Line would extend to RBS, instead of terminating in the middle of nowhere and forcing a transfer to a less-frequent, less-reliable bus.

      Forcing more bus lanes is generally a good thing. But keep in mind whether they are pushing back protected bike lane installation, which is needed now, because climate change.

      1. I’m investigating a truncation of the 157 currently. I still have a lot of data crunching to do though. :)

      2. I completely agree actually. I mainly took the 158/159 because 1. gets me closer to where I was going than Sounder and 2. cheaper than Sounder. I’m not making the argument that the 158/159 should remain, but that since Metro wants to run a bunch of downtown Seattle peak expresses, they should make sure there is room for them all to operate reliably.

        As for the 157, I’ve always thought that when Angle Lake opened, they should have taken a look at what they could do with truncating the 157 there. Certainly it wouldn’t pay for all-day half-hourly service, which would be great (and worthwhile looking at for a restructure), but doubling frequency during peak should be doable. Like they should have at least asked riders of that route in 2016 “how would you feel if you had to switch to Link, but you then get 8 trips of this bus instead of 4 or 3, and if you didn’t have to hurry out of the office at 5 to make the last bus?”

      3. If the 157 remains peak-only regardless, I would point out that connecting to Tukwila Sounder Station is much faster than connecting to Angle Lake Station, and Sounder is much faster than Link for getting downtown.

        There isn’t much there there to Angle Lake Station itself. For those wanting to connect to the airport a transfer to the 180 should suffice. It that bus stop where the routes cross each other doesn’t exist yet, add it.

        Make the route 2-way with the saved duplicate-head and headhead hours.

      4. You’ve made this point before but you still miss the point that “Sounder is not crush loaded” is not the standard. The Kent Station parking garage is at capacity. How exactly are the commuters who don’t live near it supposed to get to Sounder? Certainly not through the non-existent or unreliable feeder lines. And if you miss Sounder you get to take the 150 or the 180 and Link and will be half an hour or more late for work.

      5. “How exactly are the commuters who don’t live near it supposed to get to Sounder?”

        This is for people who are on the 158/159, not those who can’t take them.

        The 157 I don’t know, as it serves the area northeast and north of Kent Station and it would be backtracking to go to the station.

        SeaTac is a larger destination than Angle Lake so a better truncation point. Buses would also have to backtrack to get to Angle Lake Station because there’s no road across the freeway between 188th and 216th.

        The 158/159 also serve KDM P&R, so Metro would probably keep them for that until Federal Way Link opens. The part east of Kent Station is the same as a feeder route, and the part north of KDM P&R serves the P&R, so the only part of the route that’s not either of those is between the P&R and the station. It’s probably not long enough for Metro to want to split it.

      6. @Breadbaker,

        You again miss the point. If you want more and more reliable feeder service to Kent Station, scavenge the hours from the 157, 158, and 159.

        If you want more parking availability, charge for the parking, and keep raising the cost until there is sufficient availability. If you want to raise social justice concerns about access to parking, ST has an ORCA LIFT solution. Just building more free parking makes it a lot harder to have bus feeder service.

  2. “The most likely outcome is that the 179 is truncated at Federal Way Transit Center in exchange for all day service.” And higher peak frequency, which helps mitigate reliability issues, particularly time waiting for bus.

  3. Route 63 was diverted to 8th Ave in PM a couple years ago *because* of unreliable traffic on Boren. This diversion, though wonky, theoretically saves time. Furthermore, there is no Ravenna Blvd exit on the northbound I-5 express lanes (the southbound express onramp is not reversable). So the 63, along with the 64 and 76, have to exit the express lanes in the U-district, and re-enter the mainline to reach Ravenna.

    1. You are absolutely correct. I read the map incorrectly and I apologize for that error. Living in south King County, I am not as familiar with routes in north Seattle.

      1. I agree with the commute difficulty of these stops.
        Recently changed for mornings, the 63 Southbound doesn’t take Mercer exit anymore. It diverts to Union St and it’s first stop after Northgate Mall is the convention center. It arrives more timely, but it’s much further away from it’s original 1st stop after the mall.

    1. As mentioned in the article, that was my second choice.
      I felt like splitting the 37 gave that route a better chance to be successful. Shifting riders to the C-Line at the Alaska Junction, and the 120 at Delridge also sets up the hypothetical routes for transitioning to the West Seattle light rail when it opens.
      Thank you for commenting on my post!

      1. I don’t think sending Alki folks all the way south to the Alaska Junction or transferring a couple of times just to get to town is functional for people. We are nowhere near the Alaska Junction hub. I support mass transit and use it when possible, but it is clearly not functional for Seattle’s Alki Ave residents. Especially as our density increases. They should simply 1 – Run the 37 throughout the day and 2 – make a loop around Alki Ave / Harbor Ave to downtown and back coming down Admiral.

        To add ridership they need to reduce transfers, run routes during the day ( not just commute hours), and make the buses reliable ( no more ghost buses) as well as efficient (no one is going to take a mass transit trip of an hour or more for a fifteen minute drive.

        Yes that costs money but the City of Seattle has chosen to try to force us (stick instead of carrot) onto mass transit. They should put their (our) money where their mouth is and fund true, reliable, and efficient mass transit. Until then we will see gridlock and pipe dreams as the city relies on a revolving door of consultants and administrators.

  4. Eliminating the freeway stop for the 63 means all those U-district people have to go back to the 67 which is also frequently late and always… always… always overcrowded. Did I mention it’s always overcrowded? So much so that if you catch it north of campus parkway between 4 and 5:30 you risk not getting on the bus and have to wait for another overcrowded bus.

    The 67 is also much slower.

    If you’re going to do this, then metro should introduce a local and express 67. The local that makes all the stops and ferries the student the five to ten blocks north of campus. And one that goes express until at least 65th street to get the workers out of the U-district.

    1. It is wonderful that the 67 runs every 10 minutes most of the day. But having it only every 10 minutes during the times the train is running every 6 minutes means some of the buses are much more overcrowded than others. (And you just happen to be on the one that picks up two trainloads of riders transferring from downtown, because that is the bus most riders are likely to be on.)

      The need to increase peak frequency on the 67 to every 6 minutes seems inevitable. Eventually, it will need to go to every 5 minutes in 2023, when East Link opens.

      1. More people coming from the Eastside to Northgate? But no, Link will have two lines to Northgate then. And it will have one line to Northgate in 2021. So we’re really only talking about an express for one year.

        (And given the lead time to plan and approve a new route and the fact that Metro likes to limit service changes to twice a year — plus the fact Metro has no spare buses peak hours, make a new express route unlikely.)

      2. OK, so I guess the idea is that once East Link opens, the train will be more frequent, which which will then mean the 67 has to be a lot more frequent.

        Sorry, but I don’t buy it. As I wrote elsewhere, once Link gets to Northgate, the 67 is not that special. The only people that will ride buses like the 67 are those who are going in between train stops. To be fair, since Link doesn’t have that many stops, there will a fair number of people doing that (in places like Campus Parkway, 55th, etc.). But there will be plenty of buses making that trip. Between UW and U-District, there will be a bunch (45, 48, 49, etc.). Between the U-District and 65th there will also be a lot of frequent buses (45, Roosevelt RapidRide and something resembling the 67). Between 65th and Northgate there will be far fewer buses — but I don’t think that you need five minute frequency between there. If you do — if you come even close to that — then the split I have in mind would be better than ever.

        That is because the usage pattern splits. Some are just headed to Northgate (where 5th Avenue is by far the best choice). Some need to get to Maple Leaf (where Roosevelt Avenue is the best option). Having buses on both is better — riders get to their destination much faster, and buses pick up a lot more riders per hour. If it turns out you have the kind of demand that requires five minute service on both then I would be very surprised, but thrilled. My guess is that buses run every 10 minutes up to Pinehurst and every 15 minutes up to Northgate (via 5th). That should be fine.

        The 67 is really an old-fashioned run. It is basically a “high coverage” route. It runs by as much density as possible, with little regard for how long it takes riders to actually get to their destination. It manages very good frequency, in a city that still struggles with that (just look at the thread about West Seattle). Better to force riders to spend five minutes going the wrong direction than spend twenty minutes waiting for the next bus.

        Except for much of Seattle, that time has passed. If the 67 — a bus that runs every ten minutes — is too crowded, then we can certainly afford more efficient runs. The buses will be filled if you go the fast way. Might as well go that way, even if it means running *only* every ten minutes, all day and night.

    2. From a riders per minute standpoint, the 67 is an average bus during rush hour, but an above average bus the rest of the day (and at night). It still lags behind the 44, the 8 and is only a bit ahead of the 45 (all buses that don’t go to downtown). My guess is that its popularity is due to it being the only frequent bus in the area, and the primary way to get from Northgate to the U-District/UW. Like some of those other buses, it is not very reliable. Much of that unreliability is due to all of the turns between Roosevelt Avenue and Northgate. Making matters worse, a lot of the potential ridership has a straightforward alternative: walking. The button hook means that someone could get off the bus, walk down the street and arrive before the same bus — even when there is no traffic (https://goo.gl/maps/k248F8EvBZRTqb3q7). That means that the bus is taking a very slow route.

      All of this should change when Northgate Link arrives. There is no need to have the curving button-hook part of the 67. It should be straightened out, and replace the northern part of the 73 (while the other part of the 73 should be put out of its misery). Those trying to get from Northgate to the UW will take Link. Those trying to get from Northgate Way to Link still have the truncated 41, the 75, 347 and 348. Those trying to get from Northgate Way to Northgate still have the 75, 347 and 348. Those along Roosevelt Way would have a more reliable way to the UW, along with a faster way to Link (it is faster to get to Roosevelt Station than Northgate Station). The buses avoid turns, and thus spend less time waiting for the traffic lights. Pinehurst gets a more frequent connection to the UW, Maple Leaf and Roosevelt. People in Maple Leaf have a two seat ride to Northgate, but a frequent one (using the 75, 347 and 348). People used to the 73 would have to walk a little ways, but I would still keep the 77 (truncated at Roosevelt). With the savings from the truncation, you could run the 77 a bit more often.

      1. I’ll disagree on the cause of the 67s problems. Much of the reason the 67 is late isn’t because of what happens in Northgate. It’s because of A) overcrowding in the U District and B) being through-routed with the 65. If I’m waiting at the UW HUB and the 67 is running 5-10 minutes late (a not uncommon occurrence) how is that caused by left turns in Northgate?

        The 67 is somewhat unique among UW buses. Buses like the 372 or 75 pick up virtually all of their passengers on campus and once they leave campus they’re mainly dropping people off. In contrast, the 67 keeps picking up more passengers all the way up to 65th St. That leads to delays as passengers boarding and leaving a packed bus have to push past each other at each stop along Roosevelt, at least until the bus hits Maple Leaf.

        Relatively few people take the 67 all the way to Northgate. Most are taking it between UW and the U District. Changes to the route in Northgate shouldn’t be assumed to be a panacea for the 67.

      2. The bus is often late heading south as well. Wait, I have to clarify, since the bus goes both north and south on the same run. A bus is often late heading to the U-District as well.

        Changes to the route in Northgate shouldn’t be assumed to be a panacea for the 67.

        No, but in combination with Northgate Link and other changes it would dramatically improve it. There are plenty of people who ride it from Northgate to the U-District — how else are you supposed to get there? With Northgate Link, plenty of people will ride Link to Northgate, from UW, U-District and Roosevelt. There will be riders who just hop on the train for one stop (e. g. U-District to Roosevelt). Those traveling between Roosevelt and the U-District will use the 67, but they will also use the 45, which by then should have similar frequency. Oh, and you also have the Roosevelt Rapid Ride, which will be plenty frequent, and take a lot of the load off the 67.

        Keep in mind, the 67 is *not* especially unreliable. It isn’t above the lateness threshold that Metro sets at any point of the day. Nor is it especially crowded as of the last Metro report (although things may have changed). It’s biggest weakness is its button hook, which will hopefully go away when Northgate Link gets here.

      3. The 67 has a large secondary market of intra-U-District service between 65th and UW Station. Many of those will go away when Northgate Link opens.

      4. I ride the 67 a lot off-peak and in the PM peak, and in my experience it’s much more reliable southbound (to the U District) than northbound. I usually catch it S/B at Northgate/5th or 65th/Roosevelt (rarely more than 5 minutes late), and N/B at various points in the U-district and at 65th/Roosevelt (routinely 10+ minutes late). The button-hook, while inconvenient, is short (~10% of total route length) and doesn’t seem to be the source of the schedule unreliability. I agree with Larry, I think the 65 through-route and campus congestion is the issue.

      5. My hope for route 65 is that it switch to the 45th St Viaduct when Northgate Link opens, and can get to U-District Station more quickly than it gets to its closest point to UW Station currently.

        There ought to still be one route that goes between UW Med and Children’s, but that might be accomplished just as well by having the 65 continue from U-District Station down to UW Med, rather than slogging through the campus loop. But then, yielding Montlake north of UW Station to the SOVs completely might make it harder for buses just trying to get between SR 520 and UW Station.

        And as Ross points out, the J(?) Line might replace much of the 67’s purpose, though certainly not all of it, and not by 2021.

        Arguing over which routes are most worthy of getting 6-minute headway during peak misses the point. There are several that ought to get that treatment, so let the data prioritize and keep adding service to stay ahead of demand. If every 12 minutes is not enough to meet demand for a Link-connecting route during peak, jump to 6 minutes, even if it means using shorter vehicles.

      6. And as Ross points out, the J(?) Line might replace much of the 67’s purpose, though certainly not all of it, and not by 2021.

        My point is that Link itself will replace much of the 67’s purpose. Right now, for much of its route, the 67 is not unique. Lots of buses run between Husky Stadium and Roosevelt. That will continue to be the case after the next restructure.

        However, the 67 does several things that no other bus does. One is to connect Northgate with Roosevelt, the U-District and Link. Once Link gets here, you won’t need a bus to do that. The only part of the 67 that is likely to be unique is service along Roosevelt Avenue north of 65th. Even then, during rush hour you will have the 77, running just a few blocks over. With all due respect to Maple Leaf, it won’t need the type of frequency found on the E.

      7. My hope for route 65 is that it switch to the 45th St Viaduct when Northgate Link opens, and can get to U-District Station more quickly than it gets to its closest point to UW Station currently.

        Yeah, I think the either the 65 or 75 will do that, especially since it looks like the 44 won’t be extended to U-Village any time soon.

      8. Roosevelt is a strong corridor, not a coverage area. I lived at 55th and had destinations at 65th, 80th, Northgate, Meridian Ave, 50th, 45th, 42nd, Louisa/Lynn/Roanoke, Mercer, and north and south of Mercer. I only used the 66 for some of these, but others living elsewhere along the line would use the other ones. And Roosevelt/Eastlake/Fairview is growing significantly.

        The RapidRide design studied terminating at 45th, 65th, or Northgate, and settled on 65th. This implies a 67 is necessary north of 65th, and I assume it will continue south to 45th, Campus Parkway, UW Station, or continue interlining with the 65. A lot of the U-District destinations are in between the Link stations. The denser a neighborhood, the greater the expectation of in-between service.

        Metro has positioned the 67 as the primary north-south route, and tried to kill off the 71 and reduce the 73 to peak-only. There was widespread opposition to that and Metro retreated, but Metro may still see Roosevelt as the primary nort-hsouth corridor. Variations like the Northgate Way button-hook or running partly on 5th are details.

        My experience with the 67 is it’s remarkably slow, although the Ravenna area has improved with the road diet. There’s no particular bottleneck but it just crawls along and that adds up. I don’t think the problem is the entire 65 but specifically Stevens Way and Montlake Blvd. Stevens Way is congested when UW is in session, and the Montlake congestion backs up to U-Village and even 36th.

      9. “The only part of the 67 that is likely to be unique is service along Roosevelt Avenue north of 65th. Even then, during rush hour you will have the 77, running just a few blocks over.”

        The 77 will presumably be restructured in 2021. Its only stops are at 145th, 110th, and 80th, so it can’t be used for Maple Leaf to the U-District which is the 67’s primary market.

      10. North of 65th, the 67 is almost like a “hidden express” – very few people get off between 65th and Northgate. The bottleneck, at least southbound, seems to be between about 53rd and 45th, in addition to the obvious part on Pacific. I suppose if they moved the parking lane between 50th and 45th to the right side of the street to act as a bike lane buffer and then made that lane bus-only during peak, that might work, but otherwise the 67 and future RapidRide are going to suffer every afternoon.

        Bizarrely, Campus Parkway is also incredibly slow despite being a major bus corridor, because buses seem to manage to catch every single traffic light along its 5-block length. Shouldn’t there be some sort of signal priority?

    3. OK, my apologies for looking too far ahead. I wrote about what should happen (or at least what I hope will happen) after Northgate Link. But that is still a few years into the future. I never addressed Matt’s point, which is more short term. It is possible that having the 63 skip the U-District stop would make the 67 even worse. Metro should have stop data, to support or refute that worry.

      If it turns out that it is a problem, I like his idea. A 67X could make a lot of sense. Just go through the U-District, then skip the stops between 45th and 65th. Then go on to Northgate, via 5th. You could make the same stops on 5th as the 63, which means that no existing users of the U-District to 63 lose out. The bus would get to 65th a bit faster, and get to Northgate a lot faster, which means that it could help ease the crowding on the 67 while actually saving the agency money (it would pick up more riders per hour).

    4. Yeah, I agree that the 67 to Northgate is frustratingly slow, especially since it overshoots the station and turns around back to the station. I think though, that fundamentally this is a problem that the 63 is not suited to solve. What would really be warranted is a true 67 express like Matt suggested.

      Taking 1/3 of the peak of peak 41 buses to make a 67X may work (it’ll certainly be very frequent at peak). Suppose it ran UW Station to Northgate TC, making only a stop at 45th and Roosevelt on the way (maybe throw in 65th as well, to mirror the future Link stations). That would definitely alleviate crowding on the 67, but would probably exacerbate crowding on the 41 (it is very crowded even with 4 minute peak headways). Though some people would definitely prefer to make use of the DSTT again and take Link to the 67X, it would probably still leave the 41 more crowded.

      1. Yeah, my guess is that cutting frequency on the 41 would lead to a full scale rebelion. It is bad enough that the county kicked the 41 out of the tunnel well before Link gets to Northgate. Now you are cutting frequency? Light the torches and grab the pitchforks.

  5. Excellent post. I have a few comments:

    It is a virtual certainty that the route 63 will be revised whenever Metro does the restructure that will be related to the opening of Northgate Link.

    Yes, definitely. The 63 is doing several things at the same time — all of them laudable:

    1) Provide service on 5th Avenue NE between 65th and the Northgate Transit Center.

    2) Directly connect Northgate to First Hill and Cherry Hill.

    3) Directly connect the U-District (sort of) to First Hill and Cherry Hill.

    Despite the lack of reliability, the 63 manages to get mediocre ridership per hour (for an urban bus route). There are plenty of suburban express routes that are much worse, and even some of the urban express routes are worse. This suggests that the goals are solid, but the implementation is poor. I would do the following, after Link gets to Northgate:

    1) Keep the northern part of the 63 (from 65th to Northgate), but connect it to another all-day bus (like the southern part of the 26).

    2) Straighten the 49 so that it serves First Hill (basically just keeps going on Broadway, or doglegs over to 12th). This would create a good connection from the U-District to First Hill. (The connecting from the U-District to downtown becomes less important once Link gets to the U-District).

    Northgate riders heading to First Hill would have a two seat ride to First Hill, but once Link and Madison BRT is built, service will be fast and frequent (for both). Getting to Cherry Hill would require catching the 3/4, which at least is frequent if not fast.

    All that being said, the temporary changes — especially skipping the U-District stop — make a lot of sense. It would be nice if we had stop data to see how many people use that stop (along with the stop on 8th).

    1. #3 is only partly compelling. Between the U-District and First Hill you can take any U-District bus every couple minutes to UW Station, Link to Capitol Hill Station, and the streetcar or 60 or walking to First Hill. That’s much more reliable because it’s not subject to freeway congestion. If you’re going from the U-District to Cherry Hill then the 63 is more compelling because the 3/4 is a basket case and you’d have a long hilly walk from the streetcar.

      1. Taking a three seat ride to the U-District in the manner you suggest is hardly appealing. To begin with, you have traffic getting to Husky Stadium. Then you have the hellish transfer. Then you have another deep bore tunnel station (Capitol Hill) followed by what — the streetcar? That could be another ten minute wait. Oh, and you still aren’t at Cherry Hill.

        It is no wonder that ridership on the 49 is so good. People would rather slog along in a fairly slow bus than make two really bad transfers.

        More riders will switch to Link once we get the with the U-District station, and plenty more will switch if they send another bus (like the 49) north-south in the neighborhood, so they don’t have to depend on just the streetcar. The greater Central Area (which includes Capitol Hill) desperately needs a grid.

    2. Having kept the 66 running until Roosevelt and Northgate Link opens would have helped the South Lake Union riders and given those of us living south of Northgate much faster access to downtown at all hours.

  6. All of these routes will eventually be served by, and truncated at, light rail (eventually)
    Link expansion will greatly benefit our transit network, but the wait seems so long…

  7. The 216 and 219 are much better routes than the 269. Standing at Issaquah Highlands P&R, about half of the 269s one sees in both directions are empty. Investing in the bus routes people actually use is a better idea than forcing people to use the bus route people don’t use as much.

  8. I like the idea of splitting the 37. However, I would take a different approach. The southern part is very low density — similar to western Magnolia. It is also a curving, indirect route, which again is reminiscent of western Magnolia. This southern route is only justified for coverage reasons — it should run infrequently, and connect to other routes. I would go further, and connect it to the western part of the 57 — a similar low density section (albeit not quite as low density).

    In contrast, the northern part covers a relatively dense part of West Seattle. It makes sense as an all day express bus to downtown Seattle, similar to the 125. Once the mess downtown is cleaned up, it would be a cost effective, popular bus. Whatever congestion occurs for buses should be dealt with directly (via bus lanes). It shouldn’t be that difficult, really. Right now Admiral Way splits as it approaches the West Seattle Freeway. The right lane goes towards Avalon, while the left lane goes to the freeway (https://goo.gl/maps/bd6YfmL2MBxRzH1NA). The right lane could be “Exit Only Except Transit”, which would allow a bus to pass cars before the entrance (if they are backed up that far). If the ramp is still too backed up, then the bus could head towards Avalon, and turn around there. The stop sign would have to be altered to read “No Left Turn 6-10 AM except bicycles and buses”. Better yet, add a traffic light there. Continue to prevent cars from turning left there, but allow buses to turn left there. That would put the bus in the same position as the C and the 21.

    1. I don’t ever have a reason to use Route 37, but I do have to wonder what benefit splitting the route would do. Congestion appears pretty light on the 37 South segment so I’m not sure how this benefits reliability, and the segment has only Alaska Junction commercial as a non-residential station on the route. Almost no one would appear to use Route 37 South. Truncating at SODO once the Lander Bridge is completed seems like the best choice to improve reliability. There may also be an accessibility benefit to extending the West Seattle terminus but I’m not sure where.

      1. Almost no one would appear to use Route 37 South.

        That is the benefit of splitting the route. The southern end should not have the same frequency as the northern end. The northern end should have all-day direct service to downtown. The southern end should be infrequent, and simply connect to buses that go downtown.

      2. If the split should happen, I have to wonder if Route 50 should also be split — with 50 West attached to 37 South. The ride around Alki seems terribly slow for 37 South riders and Route 50 would get them to SODO much faster.

      3. If the split should happen, I have to wonder if Route 50 should also be split — with 50 West attached to 37 South.

        It is tempting, but Admiral Way does have some apartments. And it is at least a straight shot to Alki. For many the 50 is the only way to get from Alki to California Avenue (and the Junction).

        The ride around Alki seems terribly slow for 37 South riders and Route 50 would get them to SODO much faster.

        Yes, but that is the point. The 37 south is a coverage route. It isn’t meant to connect Alki with anything, really. It is extremely low density, but without it, you are asking people to walk a very long ways to catch a bus. It should run infrequently, and connect to a bus that is likely to collect a lot more riders.

        As I wrote, it makes way more sense to connect the 37 to the west end of the 57. Except in this case, don’t go downtown. Just go to Avalon, where riders could transfer to a C, 21 or 50. Of course, if the bus is full, then you might as well send it to downtown (I just don’t think it would be full much of the day).

      4. Route 50 ends about 600 feet from where Route 37 would be split. Interlining one of the Route 37 segments with Route 50 really wouldn’t create any additional transfers to reach Alaska Junction. The only effect would be in determining what frequency each leg runs, and whether or not that would be a burden. If serving the few apartment buildings on Alki near Route 50 is important, a simple jog on 61st and 63rd (streets Route 50 uses) would remedy that.

        Route 50 is also known to have reliability problems too. Lots of that is because of SODO train crossings or West Seattle congestion. While it doesn’t make the top 5 or 10, it’s not great.

      5. OK, maybe I’m not clear as to what you are proposing. You wrote that 50 west would be attached to 37 south. I assume that to mean that after the 37 gets to Alaska Junction, it goes north on California, then west on Admiral, until it goes back to Alki — forming a loop. I assume that is what you meant.

        The whole point of 37 south is to isolate the very low density part of the 37. The idea is to then run that bus infrequently (maybe once an hour). This would be a big degradation for riders trying to get from Alki to the Junction (or anywhere else on California). They would have a very infrequent bus ride. A fairly cost efficient section (California and Admiral) would be dragged down by the very low density part of the 37. I think it makes more sense to pair the low density part of the 37 with the low density part of the 57 (which is just about all of it).

        The 50 is flawed, but not in West Seattle. It connects various parts of West Seattle to each other, and to frequent buses headed downtown (like the C/21/120/125). The problem is that the bus spends too much time going to SoDo. This may be popular at the moment (because buses downtown are a mess) but won’t be in a few months. The 50 could save some time by avoiding SoDo and just continuing its run.

        Meanwhile, I would add another all-day frequent express from West Seattle to downtown — the north part of the 37. This part of Alki Avenue has plenty of apartments. Riders there should have a fast way to get downtown in the middle of the day, instead of being dependent on shuttles to the ferry, or infrequent three seat rides to downtown.

        Keep in mind, all of this is pie-in-the-sky planning. What I’m proposing is not revenue neutral. But it is quite possible that Seattle will spend additional money on transit, and that could result in better service for West Seattle. We should think ahead in terms of what bus routes are the best value. The northern part of the 37 makes sense as an all-day bus route. The 128 could be extended just a little bit to the ferry dock. You could then kill off the 773 and 775 (although you might want to run them during rush hour — when taking the ferry makes sense). Add infrequent but all-day coverage runs for the 37 south and much of the 57. Have the 50 skip SoDo and you are pretty much done.

      6. I was merely suggesting continuing Route 50 west (meaning a split at SODO) from its current Alki terminus south to then replace Route 37 south and ending back around at Alaska Junction. I’m not suggesting any relocation of the Admiral/Alki leg of Route 50.

      7. Ah, OK. Except I’m not too fond of that idea either. The big problem is the mix of demand along the route. The 50 travels through fairly densely populated neighborhoods. It runs every 20 minutes. Extending it to serve that area would cost a lot more, and not get many riders.

        Matching demand along a route is not easy. Sometimes you don’t worry about it (there are low demand places that happen to be “on the way”). Other times, you have an extension that is lower demand, but should be served by that route — just not as often. The 3 to Madrona is like that. I could see taking that approach for the southern end of the 37, although I think it is simpler to just create a circulator by combining it with the 57 and running it (and the other part of the 37) all day.

      8. There are multiple paradigms for an “unbalanced” route (i.e., higher ridership in one half).

        A) Split the route.

        B) Split two routes and create two new routes, one with the two high-ridership segments and one with the two low-ridership segments. Metro did this with the 1 and 14 (originally two routes serving 1/36 and 47/14; now three routes serving 1/14, 36, and 47).

        C) Attach a coverage segment to a stronger route. This can “save” the segment from being deleted, and potentially increase its ridership. The 62’s NE 65th segment is like that, inasmuch as it saves previous NE 55th service. Metro’s 2025 plan has some ideas to restore deleted 25 coverage this way.

        The 50 is clearly three or more routes in one. Alki-WSJ is strongest and the primary way to the beach. WSJ-SODO is questionable but Metro promotes it as a Link connection. SODO-Columbia City is “Meh, I wish it ran express between West Seattle and Beacon Hill”. Columbia City-Seward Park and Seward Park-Othello are two coverage segments, but important if we want Link feeders. Stringing them all together minimizes the number of buses and layovers, and allows epic one-seat rides. (There are probably Seward Park-Alki riders, because how else could you get there on transit?)

  9. It seems to me that there should be some way of combining the on time % of a route, how many passengers it serves a day, how many buses run a day.

    Take route 37. There are 4 buses in each direction each day and it serves 200 people daily. Based just on that, I’m honestly not sure why I should really care about the bus and how to improve it.

    Now if there was an all day route that had that sort of reliability trouble, I’d say it should be a much higher priority to fix. From Metro’s data, some routes that have all day reliability issues include the 5, 62, and 148.

    1. Good point. There are a lot of express buses that are unreliable, but run only a handful of times each day. In general the express buses are less reliable (which makes sense since there is more traffic). I think the 5 is surprisingly unreliable. It is unreliable all day, and into the evening.

    2. Hi Larry,

      The goal of this article was to look at the least reliable buses and then seek to improve them. The lack of punctuality of the bus route determined its inclusion not the number of passengers. With luck, greater punctuality would induce higher ridership. The truth of the matter is that the 37 will be truncated in West Seattle when Link opens. It will be interesting to see how Metro uses the hours saved to improve service.

      -Arthur

      1. I kind of doubt any reliability improvement will do much to bump ridership. It’s useful for coverage on the south portion, but the walkshed there is REALLY small. And for the north part of it the water taxi and the free water taxi shuttle provide more runs and faster time to downtown along the same route. The only people it really seems significantly more useful for on the north route are people wanting the slightly cheaper metro fare, or who want to get off in sodo (or for occasional times when the water taxi is cancelled for some reason).

  10. My last few shake-ups with Metro in early to mid 1990’s, seem to remember some Route 107 and 106 routing between the Tunnel and CBD and several neighborhoods.

    Which were terrific coach-handling practice for the then- brand-new Breda fleet, learning to maneuver those monsters through neighborhoods above Renton, where 30-footer would’ve been a squeeze.

    Not advocating either repeating or duplicating. But really liked driving the stretch along the south shore of Lake Washington itself. Renton CBD to Rainier Beach Station. Would’ve made a spectacular south end to to actual Metro plans to wire the Route 7 through the Tunnel, ramped via I-90 to Rainier Avenue.

    Anything like that either in service or being planned now?
    Just checking.

    Mark Dublin

    1. The 106 and 107 in the tunnel’s first decade went from the tunnel to the busway to I-5 to Swift Avenue and Myrtle-Othello, then split with the 106 on its current route to Renton and the 107 along the south Lake Washington shore to Renton. The 42 served the current 107 area in Rainier View.

      When Link started in 2009, the 106 and 107 were pulled off the freeway. The 106 west of Rainier & Henderson was restructured multiple times. The 107 west of Renton West Hill was rerouted to replace the 42’s service in Rainier View and terminated at Rainer Beach Station, later Beacon Hill Station via south Beacon. MLK north of Henderson was restructured multiple times, variously involving the 8, 38, 42, and 106.

      I don’t quite remember south Beacon (south of Myrtle). At first a 1 and 36 extension served it. Then there might have been a route from the west turning south on Beacon, but I don’t think it was the 106 or 107 which continued east to Rainier & Henderson. So there’s a gap between the 36 extension in 1990 and the 107 extension in 2009 I don’t remember.

  11. As a regular rider of the 64 (same routing as the 63 for much of the route, 41% late during PM commute) I can say that the afternoon delay is caused almost entirely by traffic on Fairview. Traffic is often at a standstill for the blocks approaching the I-5 onramp, leaving buses 20-30 minutes late. There is zero priority for transit. Hopefully Metro will consider routing the 63/64 to a parallel street.

    1. Interesting. That is exactly the area that will benefit from the Roosevelt RapidRide project. Unfortunately, it probably won’t be implemented in time to save that route. By the time it gets here, Northgate Link will already be running, and a Northgate to First Hill express is tougher to justify (although it will likely be replaced by a Lake City to First Hill express). The inability of the various agencies to actually time things properly would be funny if it wasn’t causing people so much grief (ask a bus rider in West Seattle about this).

  12. The solution to unreliable routes is to cancel them? I like your thinking. A route that doesn’t exist can’t be late.

  13. The 63 (and 64 for that matter) serves NE 45th to transition from the express lanes to mainline I-5. The southbound express lanes entrance at NE 65th is not reversible (the roadway doubles as a mainline southbound entrance), so the 63/64 cannot use the express lanes north of the U-District. Currently, the 63/64 get on I-5 using the Mercer Street express lanes, bypass 520 traffic, then get off at NE 42nd and back on I-5 for the remainder of the trip to NE 65th. The freeway stop is on the way, so it makes sense to serve it. If you want the 63 to stay on I-5, you either have to use the mainline all the way to NE 65th (likely slower than exiting and entering again) or exit at 522.

  14. I think the author is right to only recommend very minor changes to Route 63 until Northgate Link opens. We are only 25 months away — and it may happen sooner! It’s exciting that opening day is coming soon!

    1. I agree. Northgate Link can’t come soon enough imho. The U-District will finally get the light rail stop they were promised in the Sound Move plan, albeit 15 years late. Also, it’s too bad that Northgate Link got pushed back a full year (from 2020 to 2021) in the 2010 program realignment. Here’s to hoping that this extension can come online somewhat “earlier” than planned.

      1. @RossB. You aren’t kidding. One of our region’s largest employment centers with decent residential density to boot, First Hill has gotten screwed over twice now by ST.

  15. My bete noir is the 26/28/131/132. I take them a few times a month in the afternoon and evening, and they’re always 10-20 minutes late between 10am and 8pm, although they get better after that. I don’t understand why there hasn’t been a bigger outcry in Fremont over it.

    1. I’ve never understood it. They bypass the Fremont Bridge. They’re right in the middle of stadium traffic but the delays occur even on non-game days. Their North Seattle routing is on uncongested back streets. Their South Seattle routing is pretty uncongested too. The 1st Avenue South bridge could be a choke point but I’ve never seen major congestion on it. Downtown is congested but why would that affect the 131/132 more than other routes? I mostly ride them from Pike Street or Lander Street to Costco, sometimes from 45th to downtown or Costco.

      1. Hi Mike,

        If the 1st Avenue Bridge opens that can cause traffic to back up far enough to block the 131/132 from navigating the 509/99 interchange. The bridge can open for boats of sufficient size during rush hour. From the DOT website: “The bridge will open at any time for a vessel of 5,000 gross tons or more, or for a tugboat going upstream to get a vessel of 5,000 gross tons or more.”
        However, I would think this would only interfere with Northbound buses. Is it possible that the delays to northbound buses can cascade and affect the Southbound buses since they are through routed?

      2. I used to ride the 131 semi-regularly from Burien to Green Lake, and in my experience it was never any one thing, just a missed light here, slow cars there, traffic getting off in Georgetown, the time switching drivers in SoDo (I’ve had them wait up to 10 minutes there before tho), the cluster that is Downtown, traffic on the Aurora Bridge, snaking through the neighborhoods on Latona (one time a dumpster caused a 45 minute delay on the 1 lane segment there). So more death by a thousand cuts than any one thing that could be fixed. They’re very long routes with very little in the way of ‘express’ segments (509/4th Ave S is probably the closest they come for any significant length).

    2. @MikeOrr I took the 26 for almost a decade (before moving to SnoCo about 15 years ago) as the first half of my two-seat commute from Wallingford to DT and then on to parts south. Back then it was more than five minutes late SB in the morning peak maybe once or twice a week. (I boarded just a few stops before the Aurora express portion.) I found the NB routing in the late afternoon/early evening to be more unreliable so I usually took the 359 (renumbered to 358 after the Aurora Bridge tragedy) or the 360 and walked home from Aurora. Based on your comments above, it sounds like this route’s reliability has worsened in more recent years so I would have to assume that the folks in Wallingford who do rely upon this bus must be making their displeasure known to Metro. If not, then they certainly should be doing so.

    3. As a frequent 132 rider, I can vouch that the 132 is often ahead of schedule northbound (outside of peak) and actually has to dwell extra time at some stops to let the schedule catch up. This is one route where both peak and off-peak should be retimed based on the differing traffic conditions and greater time to get the larger loads on and off during peak, along with the more frequent peak stops.

      Southbound, the 132 shows up randomly, probably a result of being through-routed, just like any other outbound leg of a through-route.

      There is often a driver switch at Stadium Station northbound, which takes a few minutes. I don’t know if that is accounted for in the schedule.

      I rarely ride the 132 southbound from downtown. I much prefer to head to Beacon Hill Station and catch the mostly reliable (at least at night) 60. The 132’s southbound path around South Park is more painfully circuitous than the northbound path.

      If the 1st Ave Bridge is allowed to open during peak hours, that is also a reliability killer on the 132.

    1. Hi Jordan,

      Metro defines ontime performance as being within 1 minute early or within 5 minutes late of the scheduled time. So presumably, a bus is late if it arrives 6 minutes after it’s scheduled time.

  16. For Sammamish, I wonder if it would work well to split service at S. Sammamish P&R. You could imagine a scenario where all current Sammamish service is eliminated, and to replace it, the 554 continues at half frequency to S. Sammamish P&R from the south, and the 545 likewise continues at half frequency to S. Samamish P&R from the north. All areas that were covered with the 216, 129, & 269 get all-day service that mirrors that to Redmond and Issaquah, plus one-seat rides to Seattle for everyone all the time. Add some buffer time at Bear Creek and Issaquah Highlands to reduce unreliability, and that seems pretty solid. It splits the Sammamish unreliability by splitting the corridor into two routes.

    The losers here are people who ride from S. Sammamish to N. Sammamish, which is probably a small number of people, and would require a transfer. The big winners would be anyone going to Seattle, and people who want to ride the bus on Sunday.

    1. It seems crazy to me to run a route from Redmond, through Sammamish and Issaquah to Seattle. Is there any one actually wanting to ride the full length of that route?

      I understand Sammamish wanting commuter service to Seattle. So just truncate the line at NE 8th and let the 269 handle everything else. This is probably the most convenient turnaround while minimizing the lost service areas. The northbound buses can use the existing stop on the SE corner of NE 8th/228th, then turn right, use the roundabout at 223rd to turn around (could go all the way to 244th, if you want to add a stop out that way), then take a left back on to 228th southbound.

      People living north of NE 8th would lose direct service to Seattle, but I’m guessing that taking the 269 to the 268 would be faster for them anyway.

      1. No one wants to ride it all the way, but I could see someone living in Sammamish who may want to ride it *almost* all the way. I think it starts at Bear Creek P&R since that’s a convenient place up start, and may pick up a small number of Redmond to Issaquah commuters essentially for free. It seems like more of a problem though, since someone who parked might get on the first Seattle bus, only to find that they have to wind through Sammamish first.

        I think my proposal makes more sense than separately operating the 269 since every bus continues to Seattle in more or less the fastest way (it assumes that people in north Sammamish want to get there via 520 and people in south Sammamish want to go via I-90). The split could be moved to NE 8th, but that would be trickier because you’d need a bus layover there. Perhaps rather than an exact split, maybe some overlapping of service between S. Sammamish P&R and NE 8th makes sense as well. A problem with operating the 269 is that it has headway management inefficiencies particularly on Saturday, when it has a layover at Bear Creek for 27 minutes (with 30 minute headways). Lumping it in with a longer route with more overall frequency (even if the extension is only half frequent) adds flexibility that makes it possible to operate with less excess layover time.

      2. There is a full lane’s worth of parking on the north side of NE 8th, between 233rd and 240th that could be converted to layover space with just a couple sign posts, and a roundabout at 244th that could be used for the turnaround.

  17. Why didn’t the 8 make the cut? Is it being saved for the upcoming list of “least reliable routes in the entire country”?

    1. I have started thinking about a post-ST3 D/8 route from Ballard to Madison Park. This would complement Metro’s interest in an Uptown-Madison Park or a Smith Cove-Madison Park route. My immediate interest was getting from Capitol Hill to Elliott & Galer (Staples), but it could also help those going from mid Capitol Hill or Madison Valley to Expedia or possibly Dravus. But my concern is whether it would be too unreliable — the way the 8 already is.

      1. If the City of Seattle would make Thomas one-way westbound and John one-way eastbound in order to put curbside transit lanes on Denny Way, the Eight could become a very good east-west connector for the inner North Side.

        There would still be two auto lanes on Denny between Fairview and Broad but turns would be forbidden onto or off of Denny between those end points. The center lanes would be exclusively for “through” traffic headed to and from Capitol Hill while John and Thomas would serve as “collector / distributors” for the neighborhood. Broad Street, which leads nowhere now is wide enough for two lanes in each direction to make it the arterial between the John / Thomas couplet and Denny west of there, and Fairview is five lanes wide between Denny and Thomas. Parking is already embargoed there.

        In order to make this work the buses would have to be in a short section of bus tunnel from between Broad and Fourth Avenues to a pair of portals in the John Street ROW. West (and “north”) bound buses would exit just east of Western with a transit signal allowing unimpeded egress. Eastbound buses would operate in a descending east-side “contra-flow” lane on Elliott and then turn into a portal in the little parklet between Elliott and Western.

        John is one-way westbound in the final block and a single northside curb lane could be maintained with a right turn-only at Western for access to the buildings along the street. The tunnel portal would be between Western and the service entrance of the brown building on the south side of the street.

        There is enough of a rise between Western and Elliott and First and John that the tunnel could descend another story or so before Queen Anne North and then diagonal under the block between First North and Queen Anne North and Denny and John.

        There would be a “station” (e.g. “stop” underground between First and Second beneath Denny) for transfers from the Queen Anne Hill trolley lines.

        The bus would continue on Elliott at least as far as Expedia and preferably become Magnolia’s “Rapid Ride”.

        I grant that this does not “fix” the section between the Denny/Olive intersection and Denny and Fairview, but nothing short of an elevated busway will. This at least would greatly improve the speed and reliability of the line through the heart of SLU.

        And it’s something that the City of Seattle could pay for with the Monorail Tax because it’s clearly “Not Light Rail”.

        One complication is that in the block between Westlake and Terry the streetcar as the southern lane. That would reduce the “collector / distributor” function of Thomas to a single lane in that block. Traffic analysis would have to be completed.

  18. I am personally of the opinion that route 37 shouldn’t even go downtown at all. It already passes right by the water taxi, only to pass it right on by, go all the way around, and sit in downtown traffic.

    Instead, I would opt to replace route 37 altogether with expanded service on the water taxi shuttles. Run both of the routes year-round, and extend the Alki shuttle down beach drive during peak commute hours to avoid loss of coverage. Terminate the peak extension and Lincoln Park, since nobody from Alaska Junction is going to ride that bus all the way around – they can just get on the C-line instead.

    1. Like so much of our system, it is backwards. The only time it makes sense to run the ferry is when:

      A) It is running often.
      B) There is a lot of traffic.

      This happens only in the morning and evening. Yet that is the only time they run the express buses. They should have the shuttles run in the morning, while the express buses to downtown run the rest of the day. Of course this would cost money — but it is the type of restructure that would get decent ridership (especially if they split the 37 as noted). (See the last paragraph on this comment: https://seattletransitblog.com/2019/08/19/metros-least-reliable-routes/#comment-829120).

    2. Very long-term: Extend the 128 to Alki Beach, as has been proposed in the past, and continue on around back to Delridge Station. Riders along the coast would have frequent options in either direction to a Link station.

      I can’t speak to whether the ridership demand for the water taxi exists all day. I would be interested in comparing its carbon footprint to that of the 37, and see if one is much larger than the other. Climate change might just trump rider mode preference.

      1. I can’t speak to whether the ridership demand for the water taxi exists all day. I would be interested in comparing its carbon footprint to that of the 37, and see if one is much larger than the other.

        Generally speaking, the key is ridership. The vehicle that is full is cheaper to operate, and more energy efficient. A big train is more energy efficient than a bus, but only when it is full. Otherwise, it uses less energy to send the tiny bus around. (Fuel source matters too, obviously, but I’m talking about the amount of energy per person, not CO2 emissions).

        The thing is, the only time that boat has any chance of being full is at rush hour. Meanwhile, it is highly dependent on buses to enable its ridership (or worse yet, drivers). It isn’t like North Vancouver, which is an urban area, or the Staten Island Ferry, which connects to a rail line. Thus the same sort of vehicle (a bus) will determine whether the ferry is empty or not. If you have a lot of full buses feeding a ferry, then you might save energy. But since you don’t, you won’t.

        You only have two feeder buses — the 773 and 775. An all day 37 would pick up most of those riders. Extending the 128 to Seacrest would likely add *more* riders than are shuttled today. Except that instead of feeding a half-empty ferry, they would fill up the 37. If the 37 turns out to be overflowing, then run it more than 30 minutes a day. Again, the only time that it would make sense to run the ferry is when demand is extremely high, and the buses are stuck in traffic. That only happens at rush hour.

        Of course, what I’m talking about would likely require a second ferry. Instead of running it every 35 minutes during rush hour, you run it every 20. But then instead of running it every hour in the middle of the day, you just run buses. The cost of the ferry might be more than you save in ferry service time, but eventually it would pay for itself.

      2. A Water Taxi employee told me earlier this year that the Water Taxi is the highest revenue route, beating tje buses. Of course they don’t make a profit, but the operating loses are significantly lower. Just what I was told. No proof.

      3. @Russ, I heard several years ago that a couple bus routes like the 7 actually make a small profit?

      4. “A Water Taxi employee told me earlier this year that the Water Taxi is the highest revenue route, beating tje buses. Of course they don’t make a profit, but the operating loses are significantly lower. Just what I was told. No proof.”

        Water taxi costs per rider are on the order of $9 to Vashon and $12 to West Seattle. The Metro average is about $4, so the operating losses on the water taxi are much higher.

      5. Considering all the problems the west Seattle buses have downtown at all times of day, I’m weary of eliminating all day service on the water taxi at least until that’s resolved. For better or worse, it gets you downtown in a way that completely avoids downtown traffic.

      6. The water taxi is very valuable and I use it during the day and on weekends because the 37 only runs for 9-5 commuters. The water taxi shuts down during the day on winter schedule leaving parts of Alki with no transit options during the day. Why do we only speak of commuters? Many of us have off schedule jobs, retired, or are just not working and need to get around town like everybody else. Transit should not be profitable. It’s a public service. Why does the City of Seattle not sponsor feeder buses through neighborhoods to Metro lines?

      7. Russ:
        They do. There are Via shuttles in the Rainier Valley to connect people to Link. This blog has written about it.

      8. Thanks Pat, although municipal feeder buses (DART style) could canvas neighborhoods to bring riders to mainline Metro Buses. I think what you’re referring to is King County Metro’s support to carry riders to Sound Transit (Link stations). As for the Water Taxi, when it shuts down during the day on the winter schedule there are no transit options for Alki/Harbor residents. Once people are turned away from mass transit it’s difficult to win them back. As for costs and CO2 per person being a noble metric, if the system doesn’t meet our needs, we will not use it, which in turn decreases ridership and increases costs and CO2 per person that does use transit. Maybe aggregating those financial costs and CO2 benefit across a well run system instead of reducing it to a head count.

      9. The water taxi is very valuable and I use it during the day and on weekends because the 37 only runs for 9-5 commuters.

        That is my point. It is crazy that the water taxi runs all day, but the 37 doesn’t. The 37 (or at least a truncated version of the 37 as Arthur suggested) should run all day. It is the water taxi that should run infrequently. As it turns out, that is what happens in the winter, making the current setup really bad.

        By the way, I’m OK with running it more often in the summer, especially if it actually makes money. But folks should have an alternative way to get downtown.

        Considering all the problems the west Seattle buses have downtown at all times of day, I’m weary of eliminating all day service on the water taxi at least until that’s resolved.

        As I wrote, the system is backwards. You can get the ferry during at noon (in the summer anyway) but you can’t catch a direct bus. You can catch a direct bus — but only when downtown is most congested.

        Anyway, by the time Metro even started talking about improving the situation, the West Seattle buses will be able to get downtown faster than ever. I’m not even sure where the funding will come from. I’m just saying that if we get a bit more money in a couple years then it makes sense to have an all-day express serving Alki/Harbor Avenue, along with all-day infrequent service on Beach Drive (and similar areas).

      10. The water taxi is debatable in ordinary times, but it’s necessary during waterfront construction. It’s was the established alternative at the time, so it and its feeders are being emphasized now. The same thing happened with the E during viaduct construction. It was the existing route so it got the extra runs.

  19. I have to ask how the bus 8 was not included. During any rush hour traffic, one can easily walk past 3-4 buses parked along Denny Way with all of the other SOV idiots trying to make their way to I-5. The bus #8 is CONSTANTLY late – how did it not make the cut?

    1. The author chose the routes with the worst on-time reliability. Because these are more peak-hour routes and congestion is worst at peak hours, and long routes take more time to run their entire length, they have the worst performance in terms of percent unreliable.

      If routes were ranked by total unreliable day-long bus hours or passenger hours, I’m sure the top 5 list would be different.

  20. How important is lateness in the afternoon for outbound trips? The consequences of being late headed home are frustrating, but not the same as arriving to work late. This reliability discrepancy roughly matches the AM commute by car as well. And I believe this also matches the Service Guidelines performance measures, which allow for higher thresholds of lateness during the PM peak.

    Metro may be better off focusing on AM on time arrival, despite the higher figures for the PM peak.

    I wonder if the rider/non-rider survey has any data on customer perceptions for being late in the morning, vs the afternoon. It would help to have some customer-focused data to back this up.

    1. It’s not just the passengers arriving home late. A late outbound bus, delayed enough, will result in that bus starting the next trip late. Factor in required driver breaks and the delays will compound through the PM peak.

      AM tends to be more predictable since most travel is commute to work. In the PM you also have people going out to events and running other errands added to the commute to home.

    2. The AM peak is mainly just commuters. The PM peak is when commuters, tourists, people shopping after work, people going to evening activities, and people coming from afternoon appointments are all traveling simultaneously. So it’s harder to keep on time then.

  21. A lot of the comments here have mentioned routes that are frustratingly unreliable, even if they didn’t make the top four. Many of those routes have the same thing in common: They are just too long. The 67 gets delayed because it is paired with the 65. The 26 and 28 are paired with the 131 and 132. I don’t think it is a coincidence that people mentioned those routes, and they happen to have pairs.

    Linking together routes can save a lot of money, and in some cases, is convenient for riders. I believe the C and D were once paired, and I’m sure more than one rider took a bus from West Seattle to the Seattle Center. While it cost Metro money to split the routes and it inconvenienced some riders, I think it was worth it. Hopefully in the future we can find the money to do that with more routes. Of course chipping away at the various delays may be a better way to solve the problem.

    1. I’ve long been a supporter of splitting long routes for reliability, and applauded splitting the 174 (now 124+A) and 48 (now 48+45) and 7 (now 7+49). I used to feel the same about the 26/28/131/132, but over the years I’ve gradually come to see the benefit of a North Seattle/South Seattle route, and my limit for an interline is now the entire city rather than five miles. So I don’t think the 26/28/131/132 and 65/67 need to be split but rather they need spot transit-priority improvements.

      Another alternative is to split them but have both routes go all the way through downtown to the far neighborhood (or analagous through the U-District). That maximizes the usefulness of both routes, serves both long-distance and intra-downtown trips, and improves service to the downtown-adjacent neighborhoods which have historically been underserved.

    2. I’ve also discovered the benefit of the 7/49 interline evenings, going from southwest Capitol Hill to Little Saigon. The two-seat alternatives aren’t that great, either bus-bus or walk-Link-bus or walk-streetcar. Center City has a general transit hole going northwest-southeast. (I.e., from Uptown/SLU/Pike-Pine to First Hill/Cherry Hill/Little Saigon.) The mythical “Metro 8” line could serve it, depending on its alignment. (I.e., would it serve Swedish Broadway, Swedish Cherry Hill, or Garfield High School? Those are three different corridors.)

Comments are closed.